atomrdf_parser.inc in Feeds Atom 7
Same filename and directory in other branches
Contains the atomrd parsing functions.
File
libraries/atomrdf_parser.incView source
<?php
/**
* @file
* Contains the atomrd parsing functions.
*/
/**
* Parse the feed into a data structure.
*
* @param $feed
* The feed object (contains the URL or the parsed XML structure.
* @return
* stdClass The structured datas extracted from the feed.
*/
function atomrdf_parser_parse($string) {
@($xml = simplexml_load_string($string, NULL, LIBXML_NOERROR | LIBXML_NOWARNING | LIBXML_NOCDATA));
// Got a malformed XML.
if ($xml === FALSE || is_null($xml)) {
return FALSE;
}
$feed_type = atomrdf_format_detect($xml);
if ($feed_type == "atom1.0") {
return atomrdf_parse($xml, $string);
}
return FALSE;
}
/**
* Determine the feed format of a SimpleXML parsed object structure.
*
* @todo This whole function coudl stand to be cleaned up.
* @param $xml
* SimpleXML-preprocessed feed.
*
* @return
* The feed format short description or FALSE if not compatible.
*/
function atomrdf_format_detect($xml) {
if (!is_object($xml)) {
return FALSE;
}
$attr = $xml
->attributes();
$type = drupal_strtolower($xml
->getName());
// If there are deleted-entry elements, then we know it's not just an Atom feed
// but one with the Tombstone extension.
if (in_array(FEEDS_ATOM_TOMBSTONE_NAMESPACE, $xml
->getDocNamespaces())) {
$deleted_entry = $xml
->xpath('at:deleted-entry');
if ($type == 'feed' && !empty($deleted_entry)) {
return 'atom1.0';
}
}
elseif ($type == 'feed' && !empty($xml->entry)) {
return "atom1.0";
}
elseif ($type == "rss" && $attr["version"] == "2.0") {
return "RSS2.0";
}
elseif ($type == "rdf" && isset($xml->channel)) {
return "RDF";
}
elseif ($type == "rss" && $attr["version"] == "0.91") {
return "RSS0.91";
}
elseif ($type == "rss" && $attr["version"] == "0.92") {
return "RSS0.92";
}
return FALSE;
}
/**
* Parse AtomRDF format - for example:
*
* <entry>
* <title>Week in DC Tech: April 12th Edition</title>
* <id>19</id>
* <updated>2010-04-29T12:02:55-07:00</updated>
* <published>2010-04-12T12:16:31-07:00</published>
* <author><name></name></author>
* <content type="application/rdf+xml" xml:lang="en">
* <RDF>
* <entity>
* <title>Week in DC Tech: April 12th Edition </title>
* <properties>
* <created>2010-04-12T12:16:31-07:00</created>
* <changed>2010-04-29T12:02:55-07:00</changed>
* <uid>0</uid>
* <status>1</status>
* <promote>1</promote>
* </properties>
* </entity>
* <field type="number_float" name="field_superlative">
* <field-instance>
* <column name="value">1234.57</column>
* </field-instance>
* </field>
* </RDF>
* </content>
* </entry>
*
*/
function atomrdf_parse($feed_XML, $feed_raw) {
// declare normalized view of feed that will be returned when this method is done
$parsed_source = array();
// valid base?
$base = (string) array_shift($feed_XML
->xpath("@base"));
if (!valid_url($base, TRUE)) {
$base = FALSE;
}
/*
// Unused right now:
// Declare some canonical standard prefixes for well-known namespaces:
static $canonical_namespaces = array(
"content" => "http://purl.org/rss/1.0/modules/content/",
"dc" => "http://purl.org/dc/elements/1.1/",
'rdf' => 'http://www.w3.org/1999/02/22-rdf-syntax-ns#',
'rdfs' => 'http://www.w3.org/2000/01/rdf-schema#',
'xsi' => 'http://www.w3.org/2001/XMLSchema-instance#',
'xsd' => 'http://www.w3.org/2001/XMLSchema#',
'owl' => 'http://www.w3.org/2002/07/owl#',
'dcterms' => 'http://purl.org/dc/terms/',
'dcmitype' => 'http://purl.org/dc/dcmitype/',
'foaf' => 'http://xmlns.com/foaf/0.1/',
'rss' => 'http://purl.org/rss/1.0/',
);
// Also get any other namespaces declared in the feed
$namespaces = version_compare(phpversion(), '5.1.2', '<') ? array() : $feed_XML->getNamespaces(TRUE);
*/
// feed metadata
if (TRUE) {
// feed title
$parsed_source['title'] = isset($feed_XML->title) ? _parser_common_syndication_title("{$feed_XML->title}") : "";
// feed description
$parsed_source['description'] = isset($feed_XML->subtitle) ? "{$feed_XML->subtitle}" : "";
// feed link
$parsed_source['link'] = _parser_common_syndication_link($feed_XML->link);
if (valid_url($parsed_source['link']) && !valid_url($parsed_source['link'], TRUE) && !empty($base)) {
$parsed_source['link'] = $base . $parsed_source['link'];
}
}
$parsed_source['items'] = array();
// Parse out each entry in the feed.
foreach ($feed_XML->entry as $news) {
// Sometimes the id is the URL
$original_url = NULL;
$original_author = '';
$original_author_email = '';
$original_author_uri = '';
$guid = !empty($news->id) ? "{$news->id}" : NULL;
if (valid_url($guid, TRUE)) {
$original_url = $guid;
}
/*
// Unused right now:
// Pull out RDF statements and normalize the prefixes
// Actually there's no point in doing this since AtomRDF keeps RDF in a separate bucket.
$rdf_data = array();
foreach ($namespaces as $ns => $ns_uri) {
foreach ($rss_item->attributes($ns_uri) as $attr_name => $attr_value) {
$ns_prefix = ($ns_prefix = array_search($ns_uri, $canonical_namespaces)) ? $ns_prefix : $ns;
$rdf_data[$ns_prefix .':'. $attr_name][] = (string)$attr_value;
}
foreach ($rss_item->children($ns_uri) as $rss_property) {
$ns_prefix = ($ns_prefix = array_search($ns_uri, $canonical_namespaces)) ? $ns_prefix : $ns;
$rdf_data[$ns_prefix .':'. $rss_property->getName()][] = (string)$rss_property;
}
}
*/
// taxonomies
$additional_taxonomies = array();
if (isset($news->category)) {
$additional_taxonomies['ATOM Categories'] = array();
$additional_taxonomies['ATOM Domains'] = array();
foreach ($news->category as $category) {
if (isset($category['scheme'])) {
$domain = "{$category['scheme']}";
if (!empty($domain)) {
if (!isset($additional_taxonomies['ATOM Domains'][$domain])) {
$additional_taxonomies['ATOM Domains'][$domain] = array();
}
$additional_taxonomies['ATOM Domains'][$domain][] = count($additional_taxonomies['ATOM Categories']) - 1;
}
}
$additional_taxonomies['ATOM Categories'][] = "{$category['term']}";
}
}
// title
$title = "{$news->title}";
// body
$body = '';
if (!empty($news->content)) {
foreach ($news->content
->children() as $child) {
$body .= $child
->asXML();
}
$body .= "{$news->content}";
}
elseif (!empty($news->summary)) {
foreach ($news->summary
->children() as $child) {
$body .= $child
->asXML();
}
$body .= "{$news->summary}";
}
// author
$author_found = FALSE;
if (!empty($news->source->author->name)) {
$original_author = "{$news->source->author->name}";
if (!empty($news->source->author->email)) {
$original_author_email = "{$news->source->author->email}";
}
if (!empty($news->source->author->uri)) {
$original_author_uri = "{$news->source->author->uri}";
}
$author_found = TRUE;
}
elseif (!empty($news->author->name)) {
$original_author = "{$news->author->name}";
if (!empty($news->author->email)) {
$original_author_email = "{$news->author->email}";
}
if (!empty($news->author->uri)) {
$original_author_uri = "{$news->author->uri}";
}
$author_found = TRUE;
}
if (!empty($feed_XML->author->name) && !$author_found) {
$original_author = "{$feed_XML->author->name}";
if (!empty($feed_XML->author->email)) {
$original_author_email = "{$feed_XML->author->email}";
}
if (!empty($feed_XML->author->uri)) {
$original_author_uri = "{$feed_XML->author->uri}";
}
}
// some src elements in some valid atom feeds contained no urls at all
$original_url = _parser_common_syndication_link($news->link);
if (!$original_url && !empty($news->content['src'])) {
if (valid_url("{$news->content['src']}", TRUE)) {
$original_url = "{$news->content['src']}";
}
}
// Parse out atom values and store them
$item = array();
$item['guid'] = !empty($guid) ? $guid : $item['url'];
$item['url'] = trim($original_url);
if (valid_url($item['url']) && !valid_url($item['url'], TRUE) && !empty($base)) {
$item['url'] = $base . $item['url'];
}
$item['title'] = _parser_common_syndication_title($title, $body);
$item['description'] = $body;
$item['author_name'] = $original_author;
$item['author_email'] = $original_author_email;
$item['author_uri'] = $original_author_uri;
$item['timestamp'] = _parser_common_syndication_parse_date(isset($news->published) ? "{$news->published}" : "{$news->issued}");
$item['tags'] = isset($additional_taxonomies['ATOM Categories']) ? $additional_taxonomies['ATOM Categories'] : array();
$item['domains'] = isset($additional_taxonomies['ATOM Domains']) ? $additional_taxonomies['ATOM Domains'] : array();
// Parse out RDF values
$rdf = $news->content ? $news->content : $news;
$rdfns1 = $rdf
->children("http://www.w3.org/1999/02/22-rdf-syntax-ns#");
if ($rdfns1) {
$rdfns2 = $rdfns1
->children("http://drupal.org/");
if ($rdfns2) {
$item['rdf'] = atomrdf_custom_copy($rdfns2);
}
}
// Store this item
$parsed_source['items'][] = $item;
}
// Add any deleted items to the parsed data. These items follow the proposed
// Atom Tombstone specification: http://tools.ietf.org/html/draft-snell-atompub-tombstones-06
// @todo Ignore deleted-entry items that have a newer entry item.
//$parsed_source['deleted'] = array();
if (in_array(FEEDS_ATOM_TOMBSTONE_NAMESPACE, $feed_XML
->getDocNamespaces())) {
foreach ($feed_XML
->xpath('at:deleted-entry') as $deleted) {
$item = array(
'guid' => "{$deleted['ref']}",
'deleted' => TRUE,
'deleted_when' => "{$deleted['when']}",
'deleted_by' => array(
'name' => '',
'email' => '',
),
);
// We really should also be parsing out the by, name, and email information, too.
// However, because the XML namespace changes for those sub elements it is
// ridiculously hard to parse out in PHP, due to PHP sucking. Since they're
// not required parts of the tombstone definition, and not needed for just
// delete-and-forget processing, we're going to ignore them.
// @todo When we figure out how to parse that XML in the insanity that is
// PHP, add that in here as well.
$parsed_source['items'][] = $item;
//$parsed_source['deleted']["{$deleted['ref']}"] = $record;
}
}
// Return list of items
return $parsed_source;
}
/*
* Parse values from our custom format.
* The format consists of an 1) array of properties and 2) an array of cck properties
* @todo detect if a field_* property is a date or a sub-array and parse it out properly
*/
function atomrdf_custom_copy($xml) {
$blob = array();
foreach ($xml
->children() as $xml_child) {
$key = $xml_child
->getName();
if ($key == "field") {
foreach ($xml_child
->children() as $language) {
$language_attributes = $language
->attributes();
if (!empty($language_attributes['name'])) {
// The name 'field-instance' has a dash in it! So we are trying to get it without referring to it.
// @todo This can easily be done with the xpath() method, or with $xml_child->{'field-instance'}
$i = 0;
$values = array();
// Each field instance is a value of possible multi values
foreach ($language
->children() as $field_instance) {
$values[$i] = array();
// each column makes up information for the value
foreach ($field_instance
->children() as $column) {
$attributes = $column
->attributes();
if (!empty($attributes['name'])) {
$name = (string) $attributes['name'];
$value = (string) $column;
if (!empty($attributes['serialize'])) {
$value = unserialize($value);
}
$values[$i][$name] = $value;
}
}
$i++;
}
$key = "{$xml_child['name']}";
$blob[$key][(string) $language_attributes['name']] = $values;
}
}
$blob[$key]['#attributes'] = array(
'type' => (string) $xml_child['type'],
);
}
elseif ($key == "properties") {
// Save properties from the properties block
foreach ($xml_child
->children() as $property) {
$key = "{$property->getName()}";
$value = "{$property}";
$blob[$key] = $value;
}
}
elseif ($key == 'taxonomy') {
foreach ($xml_child
->children() as $term) {
$values = array();
foreach ($term as $data_type => $value) {
$values[$data_type] = "{$value}";
}
$blob[$key][] = $values;
}
}
else {
// There are some 'naked' properties that are not inside of the properties or fields block - save those.
$value = "{$xml_child}";
$blob[$key] = $value;
}
}
// Invokes hook_feeds_atom_atomrdf_parser_alter().
drupal_alter('feeds_atom_atomrdf_parser', $blob, $xml);
return $blob;
}
/*
* currently unused:
* This is a dumb bulk copy of properties from our rdf blob.
*/
function atomrdf_bulk_copy($xml) {
if ($xml) {
$count = 0;
$valid_count = 0;
$blob = array();
foreach ($xml
->children() as $xml_child) {
$count = $count + 1;
$key = $xml_child
->getName();
$value = atomrdf_bulk_copy($xml_child);
if ($key) {
$blob[$key] = $value;
$valid_count = $valid_count + 1;
}
}
if ($count > 0) {
if ($valid_count > 0) {
return $blob;
}
}
else {
return "{$xml}";
}
}
return NULL;
}
/**
* Parse a date comes from a feed.
*
* @param $date_string
* The date string in various formats.
* @return
* The timestamp of the string or the current time if can't be parsed
*/
function _parser_common_syndication_parse_date($date_str) {
// PHP < 5.3 doesn't like the GMT- notation for parsing timezones.
$date_str = str_replace("GMT-", "-", $date_str);
$date_str = str_replace("GMT+", "+", $date_str);
$parsed_date = strtotime($date_str);
if ($parsed_date === FALSE || $parsed_date == -1) {
$parsed_date = _parser_common_syndication_parse_w3cdtf($date_str);
}
return $parsed_date === FALSE ? REQUEST_TIME : $parsed_date;
}
/**
* Parse the W3C date/time format, a subset of ISO 8601.
*
* PHP date parsing functions do not handle this format.
* See http://www.w3.org/TR/NOTE-datetime for more information.
* Originally from MagpieRSS (http://magpierss.sourceforge.net/).
*
* @param $date_str
* A string with a potentially W3C DTF date.
* @return
* A timestamp if parsed successfully or FALSE if not.
*/
function _parser_common_syndication_parse_w3cdtf($date_str) {
if (preg_match('/(\\d{4})-(\\d{2})-(\\d{2})T(\\d{2}):(\\d{2})(:(\\d{2}))?(?:([-+])(\\d{2}):?(\\d{2})|(Z))?/', $date_str, $match)) {
list($year, $month, $day, $hours, $minutes, $seconds) = array(
$match[1],
$match[2],
$match[3],
$match[4],
$match[5],
$match[6],
);
// Calculate the epoch for current date assuming GMT.
$epoch = gmmktime($hours, $minutes, $seconds, $month, $day, $year);
if ($match[10] != 'Z') {
// Z is zulu time, aka GMT
list($tz_mod, $tz_hour, $tz_min) = array(
$match[8],
$match[9],
$match[10],
);
// Zero out the variables.
if (!$tz_hour) {
$tz_hour = 0;
}
if (!$tz_min) {
$tz_min = 0;
}
$offset_secs = ($tz_hour * 60 + $tz_min) * 60;
// Is timezone ahead of GMT? If yes, subtract offset.
if ($tz_mod == '+') {
$offset_secs *= -1;
}
$epoch += $offset_secs;
}
return $epoch;
}
else {
return FALSE;
}
}
/**
* Extract the link that points to the original content (back to site or
* original article)
*
* @param $links
* Array of SimpleXML objects
*/
function _parser_common_syndication_link($links) {
$to_link = '';
if (count($links) > 0) {
foreach ($links as $link) {
$link = $link
->attributes();
$to_link = isset($link["href"]) ? "{$link["href"]}" : "";
if (isset($link["rel"])) {
if ("{$link["rel"]}" == 'alternate') {
break;
}
}
}
}
return $to_link;
}
/**
* Prepare raw data to be a title
*/
function _parser_common_syndication_title($title, $body = FALSE) {
if (empty($title) && !empty($body)) {
// Explode to words and use the first 3 words.
$words = preg_split("/[\\s,]+/", strip_tags($body));
$title = $words[0] . ' ' . $words[1] . ' ' . $words[2];
}
return $title;
}Functions
Name |
Description |
|---|---|
| atomrdf_bulk_copy | |
| atomrdf_custom_copy | |
| atomrdf_format_detect | Determine the feed format of a SimpleXML parsed object structure. |
| atomrdf_parse | Parse AtomRDF format - for example: |
| atomrdf_parser_parse | Parse the feed into a data structure. |
| _parser_common_syndication_link | Extract the link that points to the original content (back to site or original article) |
| _parser_common_syndication_parse_date | Parse a date comes from a feed. |
| _parser_common_syndication_parse_w3cdtf | Parse the W3C date/time format, a subset of ISO 8601. |
| _parser_common_syndication_title | Prepare raw data to be a title |